8/8/2019 ant Analysis (Smoker Edition) Final
1/13
8/8/2019 ant Analysis (Smoker Edition) Final
2/13
Definitiony A procedure for the determination of the group to which an
individual belongs, based on the characteristics of that individual.
Suppose we have measurements on p characteristics for each of a
sample of individuals. We know that each individual belongs to oneof g groups, but we do not know which. Discriminant analysis
attempts to maximize the probability of correct allocation. It differs
from cluster analysis in that we have an initial data set, the training
set, whose group allocations are known.
8/8/2019 ant Analysis (Smoker Edition) Final
3/13
8/8/2019 ant Analysis (Smoker Edition) Final
4/13
Exampley Literally, "discrimination" means recognizing a difference between
two things. If I choose to buy New model Toyota car instead of
recondition car, that's an example of discrimination on the basis of
performance.
8/8/2019 ant Analysis (Smoker Edition) Final
5/13
SSPS Requirementy The observations are a random sample
y Each predictor variable is normally distributed
y Dependent variable should be categorical or nominal
y Sample size should be > 30
y There must be at least two groups or categories,
y Independent variable may have any type of measurement i.e.(ordinal, nominal, scale)
8/8/2019 ant Analysis (Smoker Edition) Final
6/13
Objectivey We take smoke is a nominal variable indicating whether the
employee smoked or not. The other variables to be used are age,
days absent sick from work last year, self-concept score, anxiety
score and attitudes to anti-smoking at work score. The aim of theanalysis is to determine whether these variables will discriminate
between those who smoke and those who do not.
8/8/2019 ant Analysis (Smoker Edition) Final
7/13
Boxs M test results tableHo : Covariance are same
H1 : Covariance are not same
Boxs M 176.474
F Approx. 11,615
df1 15
df2 600825.3
Sig. 0
8/8/2019 ant Analysis (Smoker Edition) Final
8/13
Wilks lambday Wilks lambda indicates the significance of the discriminant function.
The below table indicates a highly significant function (p < .000).
Test of
function(s)
Wilks
Lambda Chi-square df Sig.
. .
8/8/2019 ant Analysis (Smoker Edition) Final
9/13
Eigenvalues tabley This table indicates that 80.2% relationship between actual and
predicted value
Eigenvalues
Function Eigenvalue % of variance
Cumulative
%
Canonical
correlation
1 1.806a 100 100 0.802
8/8/2019 ant Analysis (Smoker Edition) Final
10/13
CanonicalDiscriminant Function
Coefficientsy D = (.024 age) + (.080 self-concept) + (.100 anxiety) + (.012
days absent) +(.134 anti smoking score) 4.543
Function1
Age .024
Self concept score .080
Anxiety score -.100
Days absent last year -.012
Total anti-smoking policies subtest .134
(Constant) -4.543
8/8/2019 ant Analysis (Smoker Edition) Final
11/13
Classification Results(a)
y 91.8% of original grouped cases correctly classified.
smoke or
not
Predicted Group
Membership Total
non-
smoker
smoker
Original Count Non smoker 238 19 257Smoker 17 164 181
% Non smoker 92.6 7.4 100.0
Smoker 9.4 90.6 100.0
8/8/2019 ant Analysis (Smoker Edition) Final
12/13
Research PaperWe would like to predict a user of Internet banking from a non-user ofInternet banking. In this case, the dependent variable is a nominal variablewith 2 levels or categories with say 1 = User and 2 =Non-user. In this case,regression analysis is no longer appropriate. Next, we have a choice ofusing a discriminant analysis which is a parametric analysis or a logisticregression analysis which is a non-parametric analysis. The basicassumption for a discriminant analysis is that the sample comes from anormally distributed population
Whereas logistic regression is called a distribution free test where thenormality requirement is not needed. This paper will only delve into theuse of discriminant analysis as parametric tests that are much morepowerful than its non-parametric alternative
(Ramayah et al., 2004;Ramayah et al., 2006).
8/8/2019 ant Analysis (Smoker Edition) Final
13/13
Thank you